LAUSR

Background Signet ring cell carcinoma (SRCC) is recognized as an uncommon subtype of colorectal carcinoma (CRC). It showed characteristic magnetic resonance imaging (MRI) manifestations. However, the MRI features post-chemoradiotherapy (CRT) were not reported, and it is unknown whether the current tumor regression grade (TRG) system by MRI (mrTRG) is applicable to SRCC. Purpose To summarize the image features of rectal SRCC on post-CRT images corresponding to the pathology, and to determine the predicting value of mrTRG compared with TRG by pathology (pTRG). Methods We retrospectively enrolled seven patients (male: female = 3:4; mean age, 45.1 years) with biopsy-pathology proved SRCC, who underwent pre- and post-CRT MR imaging followed by surgery. An experienced gastrointestinal radiologist accessed mrTRG using a 5-point grading system by mandard standard on T2 weighted image (T2WI) and then added diffusion weighted image (DWI) in a 1-month interval. Additionally, MRI features were recorded on pre- and post-CRT images as follows: pattern (target sign) and main signal intensity of T2WI, characterized manifestation of DWI, and mean Apparent Diffusion Coefficient (ADC)values. The mrTRG and all MR image features were compared to the post-operative pathology. Results At post-CRT histology, five patients got a good response (TRG 1, n = 4; TRG 2, n = 1), one patient got a partial response, and one patient got a poor response. The accuracy of MRI predicted the pathology response by mandard standard was 14% and increased to 71.4% when added DWI. After CRT, different degrees of homogeneous high SI without enhancement representing acellular mucin were observed in all patients, and the thick-ring high SI turned into a thin-target sign in most good responders. Moreover, the tumor volume decreased or slightly increased in good responders, while it markedly increased in the partial and poor responder by 57% and 73.8%, respectively. Conclusion Homogeneous high SI on T2WI and thin target sigh on DWI were the main MRI changes of RSRCC, which was corresponding to the mucinous regression and represents for good response post-CRT. The mrTRG and tumor volume was not a reliable indicator to the pathology response. We considered that DWI should be added to T2WI to evaluate RSRCC response to CRT.